AI:SECURITY

Cybercriminals and malicious actors are increasingly using AI to spread disinformation at scale – for example, by generating fake news articles and emails, or manipulating social media posts that deceive both people and systems.

AI:SECURITY combines advanced AI technologies (such as Natural Language Processing) with cutting-edge cybersecurity research to develop intelligent systems that can detect and prevent such threats. For instance, imagine an AI model that identifies a phishing email by analyzing both its content and context.

The project both aims to defend against attacks and ensure that AI systems themselves remain secure – e.g. by protecting LLMs from being hijacked or misused by malicious actors in societal applications, such as chat bots. This is a vital step toward building a safe and secure digital society.

The project funds four PhD students under the AAU AI:X initiative.